Project Summary: Our limited ability to genetically access specific cell types within the nervous system constitutes a fundamental impediment in our efforts to probe brain function and intervene therapeutically, particularly in species lacking the well-developed genetic resources available in the mouse. Targeted payload delivery using recombinant viral approaches possesses a number of potential advantages, including anatomical specificity, ease of experimental implementation, and utility in a broad range of mammalian species. Moreover, the incorporation of endogenous cell-type-selective promoter elements has shown significant promise in targeting viral payload expression to distinct neuronal subsets. To date, however, this approach has not been pursued systematically or at sufficient scale, and the stand-alone cell-type-specificity of existing viral reagents remains relatively limited. Here, we propose to establish and validate a general strategy for the identification of cell-type-specific adeno-associated viral (AAV) drivers by exploiting recent epigenomic advances in conjunction with a novel application of single-cell transcriptome analysis. First, comparative epigenomic profiling between several major neuronal subclasses in the target tissue will be used to generate a library of over 1,000 gene regulatory elements (GREs) that is enriched for highly cell-type-restricted enhancers. This barcoded GRE library will be packaged into AAV and screened en masse in vivo for cell-type-specific activity through the use of single-cell RNA sequencing. This novel approach will allow us to simultaneously assess the specificity of individual GREs against the full complement of cell types present in the target tissue. Focusing initially on mouse cortex, we will seek to validate this approach by identifying viral drivers that are specific for six excitatory and four inhibitory neuronal subtypes. To this end, we propose the following specific aims: 1) Generation of a library of GREs enriched for neuronal cell-type-specific enhancers; 2) High-throughput screening for cell-type-restricted GRE-driven AAV reporters; and 3) Validation and in vivo characterization of newly isolated GRE-driven cell- type-restricted AAV vectors. The viral drivers developed for the ten neuronal cell types in the proposed study should prove of immediate utility for a variety of cell-type-specific applications, including monitoring neuronal activity, optogenetic and chemogenetic manipulation, axonal tracing, gene delivery, and genome editing. Moreover, it is likely that these AAV vectors will retain their cell-type-specificity across brain regions and in other mammalian species. Finally, it is our hope that the proposed studies will demonstrate the feasibility of this general approach as a portable methodology, applicable in a host of brain regions, for identification of cell- type-restricted viral drivers that will be useful in a variety mammalian species.